X-Git-Url: http://mj.ucw.cz/gitweb/?a=blobdiff_plain;ds=sidebyside;f=lib%2Furl.c;h=c4662372c99fc5afc493e83e486496a756b46890;hb=8977e6ac284666f602be2a8d954842c343854e41;hp=8475a85f378498835509117b51c5b35362f208eb;hpb=90c97fa2d0d6937544205ea9ee72779753a969e6;p=libucw.git

diff --git a/lib/url.c b/lib/url.c
index 8475a85f..c4662372 100644
--- a/lib/url.c
+++ b/lib/url.c
@@ -1,8 +1,8 @@
 /*
- *	Sherlock Library -- URL Functions
+ *	UCW Library -- URL Functions
  *
  *	(c) 1997--2004 Martin Mares <mj@ucw.cz>
- *	(c) 2001 Robert Spalek <robert@ucw.cz>
+ *	(c) 2001--2005 Robert Spalek <robert@ucw.cz>
  *
  *	This software may be freely distributed and used according to the terms
  *	of the GNU Lesser General Public License.
@@ -33,19 +33,20 @@ static byte *url_component_separators = "";
 static uns url_min_repeat_count = 0x7fffffff;
 static uns url_max_repeat_length = 0;
 
-static struct cfitem url_config[] = {
-  { "URL",				CT_SECTION,	NULL },
-  { "IgnoreSpaces",			CT_INT,		&url_ignore_spaces },
-  { "IgnoreUnderflow",			CT_INT,		&url_ignore_underflow },
-  { "ComponentSeparators",		CT_STRING,	&url_component_separators },
-  { "MinRepeatCount",			CT_INT,		&url_min_repeat_count },
-  { "MaxRepeatLength",			CT_INT,		&url_max_repeat_length },
-  { NULL,				CT_STOP,	NULL }
+static struct cf_section url_config = {
+  CF_ITEMS {
+    CF_UNS("IgnoreSpaces", &url_ignore_spaces),
+    CF_UNS("IgnoreUnderflow", &url_ignore_underflow),
+    CF_STRING("ComponentSeparators", &url_component_separators),
+    CF_UNS("MinRepeatCount", &url_min_repeat_count),
+    CF_UNS("MaxRepeatLength", &url_max_repeat_length),
+    CF_END
+  }
 };
 
 static void CONSTRUCTOR url_init_config(void)
 {
-  cf_register(url_config);
+  cf_declare_section("URL", &url_config, 0);
 }
 
 /* Escaping and de-escaping */
@@ -149,6 +150,29 @@ url_enescape(byte *s, byte *d)
   return 0;
 }
 
+int
+url_enescape_friendly(byte *src, byte *dest)
+{
+  byte *end = dest + MAX_URL_SIZE - 10;
+  while (*src)
+    {
+      if (dest >= end)
+	return URL_ERR_TOO_LONG;
+      if (*src < NCC_MAX)
+	*dest++ = NCC_CHARS[*src++];
+      else if (*src >= 0x20 && *src < 0x7f)
+	*dest++ = *src++;
+      else
+	{
+	  *dest++ = '%';
+	  *dest++ = enhex(*src >> 4);
+	  *dest++ = enhex(*src++ & 0x0f);
+	}
+    }
+  *dest = 0;
+  return 0;
+}
+
 /* Split an URL (several parts may be copied to the destination buffer) */
 
 byte *url_proto_names[URL_PROTO_MAX] = URL_PNAMES;
@@ -201,18 +225,28 @@ url_split(byte *s, struct url *u, byte *d)
     {
       if (s[1] == '/')			/* Host spec */
 	{
-	  byte *q, *w, *e;
+	  byte *q, *e;
+	  byte *at = NULL;
 	  char *ep;
 
 	  s += 2;
 	  q = d;
 	  while (*s && *s != '/' && *s != '?')	/* Copy user:passwd@host:port */
-	    *d++ = *s++;
+	    {
+	      if (*s != '@')
+		*d++ = *s;
+	      else if (!at)
+		{
+		  *d++ = 0;
+		  at = d;
+		}
+	      else			/* This shouldn't happen with sane URL's, but we need to be sure */
+		*d++ = NCC_AT;
+	      s++;
+	    }
 	  *d++ = 0;
-	  w = strchr(q, '@');
-	  if (w)			/* user:passwd present */
+	  if (at)			/* user:passwd present */
 	    {
-	      *w++ = 0;
 	      u->user = q;
 	      if (e = strchr(q, ':'))
 		{
@@ -221,8 +255,8 @@ url_split(byte *s, struct url *u, byte *d)
 		}
 	    }
 	  else
-	    w = q;
-	  e = strchr(w, ':');
+	    at = q;
+	  e = strchr(at, ':');
 	  if (e)			/* host:port present */
 	    {
 	      uns p;
@@ -233,7 +267,7 @@ url_split(byte *s, struct url *u, byte *d)
 	      else if (p)		/* Port 0 (e.g. in :/) is treated as default port */
 		u->port = p;
 	    }
-	  u->host = w;
+	  u->host = at;
 	}
     }
 
@@ -537,7 +571,7 @@ url_error(uns err)
 /* Standard cookbook recipes */
 
 int
-url_canon_split(byte *u, byte *buf1, byte *buf2, struct url *url)
+url_canon_split_rel(byte *u, byte *buf1, byte *buf2, struct url *url, struct url *base)
 {
   int err;
 
@@ -545,19 +579,19 @@ url_canon_split(byte *u, byte *buf1, byte *buf2, struct url *url)
     return err;
   if (err = url_split(buf1, url, buf2))
     return err;
-  if (err = url_normalize(url, NULL))
+  if (err = url_normalize(url, base))
     return err;
   return url_canonicalize(url);
 }
 
 int
-url_auto_canonicalize(byte *src, byte *dst)
+url_auto_canonicalize_rel(byte *src, byte *dst, struct url *base)
 {
   byte buf1[MAX_URL_SIZE], buf2[MAX_URL_SIZE], buf3[MAX_URL_SIZE];
   int err;
   struct url ur;
 
-  (void)((err = url_canon_split(src, buf1, buf2, &ur)) ||
+  (void)((err = url_canon_split_rel(src, buf1, buf2, &ur, base)) ||
    (err = url_pack(&ur, buf3)) ||
    (err = url_enescape(buf3, dst)));
   return err;